Is auditing ftruncate useful?

Steve Grubb sgrubb at redhat.com
Fri Mar 6 16:59:50 UTC 2020


On Monday, February 10, 2020 6:29:22 PM EST Casey Schaufler wrote:
> On 2/10/2020 3:05 PM, Orion Poplawski wrote:
> > On 2/10/20 3:54 PM, Paul Moore wrote:
> > So, this is all reasonable.  But why do I get this with fchown which also
> > takes a file descriptor?
> > ...
> > 
> > It's this disparity between fchown and ftruncate that caught my
> > attention.
> 
> fchown changes the security state (mode bits) of the file,
> whereas ftruncate changes the content of the file. The former
> is clearly security relevant, the latter is not.

Well, security relevant or not, the requirement that the rule meets is 
located here:

https://www.niap-ccevs.org/MMO/PP/-442-/#fau

File and object events (Successful and unsuccessful attempts to create, 
access, delete, modify, modify permissions), 

Since they separate modify and modify permissions, they clearly want changes 
to content. Now, in the real world is that necessary? Maybe only in super 
important scenarios where you have to know any change to anything.

-Steve





More information about the Linux-audit mailing list